An Approach for Identifying Cytokines Based on a Novel Ensemble Classifier

Algorithm 2

Under-sampling applies 𝐾-means clustering.

(1) Input: positive sample set 𝑆^′, negative sample set 𝐵𝑆^′≪|𝐵|;

(2) Output: the new negative set 𝐵^′𝐵^′=𝑆^′;

(3) Calculate the number of samples in two sets, 𝑘 to 𝑆^′, 𝑛 to 𝐵;

(4) Select 𝜆𝑛 (𝜆 is defined as under-sampling ratio, 0<𝜆<1) samples randomly from set 𝐵 as initial clustering centroids, 𝜆=0.2 in

  our paper;

(5) Repeat;

(6) Calculate distances (Euclidean Distance) of each sample to all the clustering centroids;

(7) Choose the nearest clustering centroids and add them to certain clusters;

(8) Find the new centroids of all the new clusters;

(9) Until each cluster stability;

(10) Define the final 𝜆𝑛 centroids as 𝐵^′;

(11) Output 𝐵^′.